Project Summary Lung cancer is the leading cause of cancer-related mortality in the United States, with figures projected to again exceed those for breast, prostate, colorectal, and pancreatic cancers combined in 2017. In the wake of the findings of the National Lung Screening Trial (NLST), the widespread implementation of lung cancer screening programs is anticipated to greatly improve outcome statistics through the early detection of lung cancer - when surgical intervention is potentially curative. Unfortunately, as many as 1 in 5 patients with stage IA disease will die from disease recurrence within 5 years of tumor resection. The long-term objective of this project is focused on the development of a new diagnostic tool that will improve our ability to prognosticate recurrence in stage IA non-small cell lung cancer (NSCLC) and serve as a means to select stage IA patients that would benefit from adjuvant treatment options or closer surviellance. The objective of this pilot project is to collect data demonstrating the power of our transcriptome-guided `targeted proteomic' approach by identifying a series of novel circulating biomarkers that have strong prognostic value in stage IA lung adenocarcinoma. Our project hypothesis contends that disease recurrence in stage I NSCLC results predominantly from occult, micro-metastatic disease, which is (mechanistically) the product a phenotypic transdifferentiation in the primary tumor. Tumor cells that undergo this `epithelial-to-mesenchymal' transition (EMT) will express a distinct set of proteins from non-metastatic tumors, including those they secrete or shed (i.e. the `secretome'). With this, our primary goal is to identify a series of circulating biomarkers for prognosticating recurrence in stage IA NSCLC using a transcriptome-guided `targeted proteomic' approach for biomarker discovery. Three specific aims are proposed: In Aim 1 we will use transcriptome data sets from cases of stage IA lung adenocarcinoma reposited by the Cancer Genome Atlas (TCGA) to identify recurrence-associated differences in genes expression. Candidate biomarkers meeting thresholds for significance (p?0.01, false-discovery rate q?0.01 and ?1.5 fold difference) will then be mined for a virtual `secretome' using established bioinformatics methods. Candidate (secretomic) biomarkers will then have their expressed products (i.e. proteins) confirmed in plasma, with the top 25 candidates targeted for quantitative, mass spectrometry-based assay development in Aim 2. Once analytically developed and validated, these assays will be used in Aim 3 to screen 200 cases of stage IA lung adenocarcinoma from our institutional repository. Performance characteristics for all biomarkers will be established and multivariable statistical methods used to formulate an optimized biomarker panel with (cross- validated) sensitivity and/or specificity values ?75% for prognosticating disease recurrence. Finally, we will confirm (preliminary validation) panel performance against 100 specimens of stage I lung adenocarcinoma from the Mayo Clinic. We envision this study will provide an important molecular tool that will help improve treatment algorithms for patients with stage I disease and augment the benefits provided by lung cancer screening.